ABSTRACT. The flying capacitor multilevel converter is positioned to become a key technology in emerging power electronics applications. However, maintaining the flying capacitor voltages balanced at their nominal values presents a major challenge for these designs and has impeded their industrial adoption. While prior literature has identified the input filter ripple as the dominant mechanism causing imbalance, a detailed analytical model of this effect has been missing. To address this research gap, this work symbolically derives and experimentally validates a new three-level converter model that captures the effect of the input ripple, enabling accurate prediction of the flying capacitor voltage imbalance.

ABSTRACT. The flying capacitor multilevel (FCML) converter has demonstrated impressive power densities and high efficiencies. In pursuit of increased power density, the capacitance value of the flying capacitors is often reduced. This work investigates the effects of flying capacitor sizing on the steady-state behavior of the inductor current and capacitor voltages. A piecewise parabolic model is developed to include flying capacitor voltage ripple in circuit operation and provide analytical insights. The capacitor voltage ripple, overlooked in standard analysis of the converter, leads to flying capacitor voltage imbalance, with the worst balancing at nominal duty cycles. Considering the flying capacitor voltage ripple, the typically used linear-ripple approximation for the inductor current is not valid. Consequently, the harmonics of the inductor current ripple appear at multiples of the switching frequency as opposed to multiples of the effective switching frequency as typically modeled. The results from the model are experimentally validated with a 4-level FCML converter prototype and show excellent matching.

ABSTRACT. The flying capacitor multilevel (FCML) converter has received renewed attention due to recent demonstrations of its superior power density and efficiency relative to existing solutions. Modulation of the switching transitions by current programming shows potential for improved natural balancing of the flying capacitor voltages but is relatively unexplored due to associated practical implementation challenges. An output voltage discontinuity in the steady-state transfer characteristic at the nominal duty ratio is identified and discussed. The discontinuity is eliminated with the introduction of an additional reference current and a modified state machine design, both of which are validated in hardware.

ABSTRACT. The challenge of balancing the flying capacitor (FC) voltage without sensing it remains a critical issue in multilevel FC converters. This digest proposes voltage balancing using digital current-mode control with emulating ramps and sampling inductor valley current twice per switching cycle. Thereafter, a current error correction logic has been incorporated for faster FC voltage balancing. The slope requirement of the emulating ramp has been determined analytically from the steady-state stability analysis. The proposed technique works satisfactorily, even during the start-up process. Along with simulation results, experimental case studies endorse the robustness of the proposed technique for a 48V to 12V, 200KHz laboratory hardware prototype.

ABSTRACT. A DC transformer (DCX) is a DC-DC converter that operates at a fixed voltage conversion ratio, allowing for optimized design under constant voltage conditions and the potential for higher performance. The Series-Bridge DCX (SB-DCX) topology has recently been used in DCX applications due to its inherent, uncontrolled DCX operation. However, a convenient and scalable time-invariant dynamic model for this topology is missing. This digest examines the challenges of developing a time-invariant dynamic model and proposes the use of half-cycle averaging that results in a linear time-invariant (LTI) DC-terminal dynamic model with a convenient circuit-based description. Simulation and hardware results are presented, demonstrating a strong correlation between the observed dynamic behavior and the predictions of the developed model.

ABSTRACT. Switch conductance regulation is used in favor over dynamic off-time modulation in resonant SC DC-DC converters due to its smaller output voltage ripple and operation at a constant frequency. However, at light load the resonant oscillations are significantly damped because of the increased Rdson of the power switches. A modeling approach is presented and experimentally verified showing that under these conditions it is beneficial to scale both Rdson and fswitch to further reduce switching losses while still maintaining a reasonably small output voltage ripple.

ABSTRACT. The digest investigates the impact of higher sampling rate on the dynamic model of wireless battery chargers. A series-series resonant converter topology is considered for analysis. An exact discrete-time (DT) model is adopted for capturing and comparing the dynamics pertaining to different sampling rates. Once per-period sampling is a natural choice for DT modeling, however, the number of considered subintervals and thus, the modeling efforts get reduced if the state variables possess some waveform symmetry. Half-wave anti-symmetry of ac variables is often leveraged in the DT model, requiring double-frequency sampling. In this work, the effect of a double sampling rate compared to once per-period sampling on the small-signal dynamic model is studied while correlating the resonant dynamics with aliased frequencies of resonant tanks.

ABSTRACT. This paper presents Chiplet-LEGO, a single-input multiple-output (SIMO) linear extendable group operated (LEGO) voltage regulator module (VRM) architecture as a new development of the LEGO-VRM family. This architecture not only offers a high voltage conversion ratio but also delivers multiple output voltages, making it well-suited for VRMs in system-on-chip (SoC) applications with multiple voltage rails. The Chiplet-LEGO architecture facilitates soft-charging and vertical power packaging and achieves high efficiency and power density. The Chiplet-LEGO scales well to numerous number of independently regulated output voltages to meet specific application requirements. In this paper, a 24 V-to-1 V 120 W dual-output Chiplet-LEGO prototype is presented.

ABSTRACT. The power density of a single-phase ac-dc converter is often determined by the dc-link capacitance, which is sized to process the double line frequency ripple power in the single-phase operation. An Active Power Decoupling (APD) can help relax the requirements for dc-link capacitance by buffering the ripple power to an auxiliary capacitance. However, implementing feedback control of such a circuit requires either intrusive current sensing or complex control schemes. This paper proposes an APD technique based on an Extremum-Seeking Controller (ESC) which is simple to implement, robust, and only involves sensing of the dc-link voltage to buffer the ripple power. Simulation and experimental results on a 6.6 kW prototype validate the effectiveness of the proposed scheme. The final paper will elaborate on the design guidelines for tuning the proposed controller and present extended experimental results.

ABSTRACT. This paper introduces a new speed sensorless MTPA control strategy tailored for solar-powered Permanent Magnet Synchronous Motors (PMSM), predominantly employed in water pumping applications. Unlike conventional vector control schemes, the proposed control strategy eliminates the need for observers and avoids complex startup routines. Instead, the proposed controller tracks the Maximum Torque per Ampere (MTPA) trajectory by a simple selection of voltage gain coefficients that can be easily implemented using a low-cost microprocessor. The proposed controller is also capable of quickly stabilizing the drive under transients, even with a DC link of low stiffness. Simulation and experimental results verify the effectiveness of the control scheme under varying operating conditions.

ABSTRACT. In this digest, we propose an integral-based maximum power point tracking (MPPT) algorithm for point absorber wave energy conversion (WEC) systems. A permanent magnet synchronous generator (PMSG) is coupled to the point absorber and its drive implements the proposed MPPT control. While the rotating frame of reference of the PMSG implements power control, the slower mechanical frequency of the incident waves are processed with an additional transformation that yields another set of dc dynamics. The computed dc power in the new reference frame, which is focused on wave dynamics, is input to the MPPT control law that tunes the emulated resistance of the machine drive to extract peak power from the wave absorber device. We derive a stability condition for the proposed controller and validate our control design on a simulated 10kW system.

ABSTRACT. This paper proposes a passivity-based control method for an IPM motor drive system to ensure stability. The proposed method is based on the design of an additional feedforward (FF) compensation loop, which makes the frequency characteristics of the DC input impedances resistive or passive. The method utilizes an extended small-signal IPM model and analytical expressions considering FF compensation. Frequency characteristics of the proposed method by the impedance method are investigated and assured to be passive. The system stability is validated analytically and experimentally.

ABSTRACT. In this paper, a new hybrid approach is developed to model the power stage of a pulse-width-modulated switching converter in continuous conduction mode. First, we derive the line transfer function for the power stage based on the basic state-space averaging model. Second, we reconstruct the line transfer function into an equivalent average circuit using the time- and transfer-constant technique. The precision of the proposed hybrid modeling approach is verified by simulation results in SIMPLIS for the traditional buck converter.

ABSTRACT. This paper introduces the implementation of electromagnetic transient (EMT) simulations of a photovoltaic (PV) inverter module using the Implicit-Explicit (ImEx) solver in the Suite of Nonlinear and Differential/Algebraic Equation Solvers (SUNDIALS). This study demonstrates the effectiveness of the ImEx solver in overcoming the challenges inherent in simulating the complex dynamics of PV inverter modules. Furthermore, using SUNDIALS' ImEx solver module ARKODE for EMT simulation automates key aspects of the process, such as numerical integration, providing substantial benefits including enhanced consistency, faster implementation, reduced human error, and the capability to handle the complexities of advanced numerical integration. By conducting comparative simulations with an implicit method used in commercial software, the research showcases the ImEx solver's capability in achieving high accuracy and reliability. Results indicate that leveraging the ImEx approach significantly enhances modeling fidelity and reduces simulation setup times, offering a promising tool for the EMT analysis of PV inverter systems in power electronics-dominated power grids.

ABSTRACT. The Robicon-type drive has remained a popular medium-voltage drive option since its conception. It is often found as a part of complex multi-converter configurations. Considering this, there is a surprising lack of literature on the average modeling of topology itself. This paper presents a first-hand effort at attempting to model the average nature of a five-level Robicon-type drive. The topology is modeled as a cascade connection combining three individual stages. The developed average-value model (AVM) is validated using simulation tools against a corresponding switching. The results corroborate the validity of the developed AVM.

ABSTRACT. Ripple-based control strategies offer superior transient performance compared to traditional current-mode and voltage-mode control schemes, making them well-suited for applications requiring fast response to large load variations. This digest presents the modeling and design of a 2MHz GaN-based constant-on-time (COT) virtual-ripple-controlled (VRC) buck converter, designed to convert a 5–12V input to a 1.2V output for high-performance computing (HPC) applications. Simulation results demonstrate that the converter supports up to 20A load current and effectively manages load transients from 1A to 20A with a 5A/µs slew rate. GaN FETs are employed for their high figure-of-merit (FoM) and compact size, contributing to improved efficiency and power density.

ABSTRACT. Electrocaloric heat pump systems provide a sustainable cooling and heating solution, but require highly efficient power electronics for cyclic charging and discharging of electrocaloric capacitors. This was achieved using multilevel converters for ideal capacitive loads. However, multilevel operating without balancing for real electrocaloric capacitors leads to drifting inner buffer voltages and eventually to inefficient 2-level operation. This work proposes a purely software-based balancing technique. A 4-level GaN-based converter is built. Compared to unbalanced operation, balancing reduced the measured power loss by 24.1%, which will increase the coefficient of performance (COP) of future electrocaloric heat pump systems. The balancing operation experimentally achieves the stabilization of the internal buffer voltages and thereby enables a steady-state multilevel operation.

ABSTRACT. Isolated converters are essential in Electric Vehicle (EV) systems for their versatility. This work presents an isolated on-board converter for EV charging, based on an interleaved Forward topology with a coupled inductor, doubling the switching frequency in passive elements. Operating in Continuous Conduction Mode (CCM) with Zero Voltage Switching (ZVS), it enhances efficiency. A State-Space Averaging (SSA) model was developed, and a robust control strategy was designed. A prototype was tested under load step conditions, demonstrating stable operation. The model accurately represents the converter's behavior in the frequency domain.

ABSTRACT. Grid-forming (GFM) controllers in photovoltaic (PV) systems encounter distinct challenges since ac power demand may exceed PV generation capacity. In this digest, we propose a multi-mode control architecture for two-stage grid-tied PV systems by introducing a limited grid-forming mode to sustain operation during capacity-limiting disturbances. The framework combines three operational modes, namely a pre-synchronization mode for startup, a mode with prototypical GFM dynamics, and the proposed limited GFM mode when power constraints are reached. The ability to ride through large disturbances requires only dc-dc controller adaptation while the dc-ac converter contiguously retains one control law. Accordingly, this approach minimizes implementation complexity. Finally, simulations validate seamless ride through during large disturbances.

ABSTRACT. This digest compares the 2N+1 and N+1 operating modes of the single-phase AC-AC MMC. Unlike conventional DC-AC energy conversion, the AC-AC MMC offers versatile operating combinations based on the number of levels in the common-mode voltage vpwm and the differential-mode output voltage vo. By applying the multilevel modulation strategies such as PS-PWM and Staircase, these voltages improve power quality at the converter's input and reduce the output voltage. The 2N+1 operating mode results in a lower voltage level across the input inductor of the AC-AC MMC converter when using the same number of cells in each arm, compared to the N+1 operating mode. Simulation results demonstrate the AC-AC MMC operating under medium-voltage and medium-power conditions, emphasizing the key differences between the two operating modes.

ABSTRACT. This paper proposes an optimized design for complete zero voltage switching (ZVS) operation of the phase-shift full-bridge DC-DC Converter with series-connected transformers. This topology extends the ZVS range at light load, compared with the standard phase-shift full-bridge Buck topology, by exploiting the energy stored in the transformer’s magnetizing inductances. However, at full load, ZVS condition still requires a suitable leakage inductance, which reduces the effective duty-cycle. A design procedure is proposed that allows a complete ZVS operation without relying on the transformer’s leakage inductance, that can be minimized. Experimental results will follow, highlighting the effectiveness of the proposed design.

ABSTRACT. This paper describes a three-phase soft-switching traction motor drive with a sinusoidal output voltage. A variable-frequency control algorithm is developed to achieve soft-switching at all operating points. This is done using current ripple generated in both the leakage and magnetizing inductances. The coupled inductor allows soft-switching, high frequency operation. The sinusoidal inverter output is filtered and has a stable common mode. It is shown that the stored energy in the coupled inductor approaches half that of interleaved inductors, and reduces further with level count. The inverter is compared to interleaved soft-switching inverter topologies. The motor drive function is verified in simulation.

ABSTRACT. Stealth cyber intrusions in networked DC microgrid (DCMG) clusters employing primary droop control can severely impact system stability while remaining undetected. While significant efforts focus on strengthening cybersecurity, understanding the direct effects of such intrusions on microgrid dynamics is equally critical. This paper analyzes the stability of parallel-connected DCMG clusters under stealth cyber intrusions of varying intensities. The intrusion is modeled as a non-linear and discontinuous input, capturing the strategic disruption of the behavior of the system by the adversary. A small-signal model of the affected DCMG cluster is developed, and stability is assessed using the quasi-linearization method. The analysis reveals that even with droop-controlled power sharing, stealth cyber intrusions can introduce instability, leading to performance degradation and potential system-wide failures. The simulation results are presented to validate the theoretical findings, highlighting the vulnerability of DCMG clusters to covert cyber threats.

ABSTRACT. This paper presents a novel two-stage optimization strategy to improve efficiency in active cell balancing for high-voltage lithium-ion battery packs. The proposed method utilizes a linear programming formulation, the Transportation Problem, to optimize charge redistribution, thereby minimizing conduction losses and balancing duration. Additionally, principles from Graph Theory, particularly Eulerian Paths, are used to reduce switching losses. In contrast to conventional rule-based controllers, the proposed algorithm achieves faster balancing, fewer switching events, and improved efficiency. The results demonstrate considerable improvements in balancing performance, reducing inefficiencies typically associated with high-voltage battery packs.

ABSTRACT. This paper proposed a new grid-forming control scheme for operating grid-interfaced PV inverters as black-start resources. Compared with the generic PV control proposed by the Western Electricity Coordinating Council (WECC) and the traditional voltage regulation control, the proposed control shows eminent advantages: i) it is applicable to handle large black-start transients, such as transmission line energizations and load pick-ups; ii)it can be simulated with great fidelity of inverter dynamic behaviors; iii)it features grid-side frequency and voltage regulation functionalities, which enhance the system stability in the power restoration process. Simulation results based on a 5-bus PV-interfaced black-start testbed are provided for verification purposes

ABSTRACT. Efficient modulation of radio-frequency (RF) power can be challenging in applications with varying load impedance. Conventional solutions often use multi-stage systems to achieve the required wide power range and load range. However, these approaches can be complex, costly, and inefficient. This digest presents an approach for designing a wide-range current-mode Class-D (CMCD) inverter as a single-circuit alternative to these multi-stage systems. The proposed topology has a simpler architecture and achieves a comparable power and load range while potentially improving overall efficiency. Key design trade-offs and considerations are also discussed.

ABSTRACT. In power converters, active device losses are distributed between switching and conduction loss, which scale in opposite directions with transistor gate width. Focusing on GaN devices, this work describes the relationship between device losses, parasitic, and size, and a simple analytical size optimization is performed to minimize the overall losses. Additionally, the results of this optimization are related to common transistor figures of merit, and an approach to discrete device selection is provided. Simulation and experimental results are presented for GaN-based 50 W buck and hybrid switched capacitor converters.

ABSTRACT. Miniaturization and performance enhancement of single-phase ac/dc power electronic converters is challenging owing to commonly employed multi-stage conversion and dc regulating buffer capacitors. The recently proposed harmonically partitioned power converter (HPPC) enables minimum energy storage sizing of the buffer capacitor, unity power factor operation, galvanic isolation, and dc output voltage regulation within a single conversion stage by leveraging bidirectional switches (BDS) and transferring power through a high frequency ac tank rather than a dc link. This work presents an analytical power loss model for the bidirectional switches employed in a single-branch HPPC which considers the time-varying phase shift between currents carried and voltages blocked by the BDSs.

ABSTRACT. This paper proposes a cost-effective wave energy harvesting system for grid-forming applications using a permanent magnet synchronous machine and a Lyapunov-based controller for torque and speed regulation. A battery-capacitor dc bus enhances reliability under low wave conditions by supporting load power delivery. On the consumer side, a similar control architecture ensures low total harmonic distortion in output voltage under unbalanced or nonlinear loading. The overall system is designed to maintain operational efficiency and reliability across varying wave conditions. System performance is validated through simulations in MATLAB/Simulink and PLECS, with additional experimental and real-time case studies reserved for the final version.

ABSTRACT. Serial-connected photovoltaic (PV) power systems are commonly used for renewable energy generation. However, partial shading on PV panels can cause significant power losses. To address this issue, differential power processing (DPP) structures can be implemented to improve the overall efficiency of the PV system. The topology used in this work is a split-inductor boost converter used to control three series-connected PV submodules. In this work, we propose a discontinuous conduction mode (DCM) switching scheme for this DPP converter architecture that allows for the current through each inductor to be different. Simulation results demonstrate that the proposed switching scheme enables the PV system to operate close to its theoretical MPP, similar to existing DPP converter architectures but with fewer components.

ABSTRACT. The enhancement-mode gallium nitride (GaN) high-electron-mobility transistor (HEMT) offers faster switching speed but introduces large dv/dt, which causes crosstalk issues and further raises the risk of false turn-on. Traditional gate drivers address this by increasing the gate resistance to reduce switching speed, but this approach also elevates switching loss, especially at high temperature because GaN HEMT’s transconductance decreases with increase of temperature. To mitigate this issue, this paper proposes a driver circuit featuring adaptive gate resistance under different temperatures. By incorporating a negative temperature coefficient (NTC) resistor, the gate resistance decreases as temperature rises. As a result, GaN devices can switch faster at high temperature, thereby reducing switching losses. Compared with the traditional driver circuit, the proposed method improves the switching speed by 31% through the adaptive gate resistance and maintains the same maximum crosstalk voltage peak under a wide temperature range.

ABSTRACT. Power converters in automotive applications generate high-frequency switching voltages and currents that act as primary noise sources, producing electric and magnetic dipoles, responsible for radiated emissions. Identifying these sources in the near-field region remains challenging. This digest models the critical near-field magnetic and electric emissions caused by the switching transients of a SiC MOSFET in a DC-DC half-bridge converter. Experimental validation is performed by measuring near-field emissions from the converter, providing insights into the dominant radiated noise sources and their type of emission(s). The full paper will extend the modeling and near-field measurement to single and three-phase power converters.

ABSTRACT. The increased popularity of the flying capacitor multilevel (FCML) converter, as well as the use of wide-bandgap devices in this topology has opened up a number of questions around conducted electromagnetic interference (EMI). Common-mode EMI is generated by parasitic capacitances throughout the circuit, and the currents arising through them as a result of large dv/dt switching events. The mechanisms by which the FCML converter generates conducted common-mode EMI is an unexplored area of research. This paper proposes modeling a lumped parasitic capacitance for each switch to understand how common-mode currents are generated in multilevel operation. This paper derives the common-mode EMI scaling behavior from first principles for an FCML converter, and provides hardware validation with a 6-level converter prototype.

ABSTRACT. A DC transformer (DCX) is a DC–DC converter that operates at a fixed voltage conversion ratio, enabling optimized design. The Series Resonance Converter (SRC) operating in Discontinuous Conduction Mode (DCM) is commonly used as uncontrolled DCX. However, it cannot support bidirectional power flow and suffers from high losses under light-load conditions. This digest explores the use of a Dual Active Bridge–SRC (DAB-SRC) topology, where both bridges are active and switch symmetrically and at near series resonance frequency to achieve ideal bidirectional DCX operation. It is observed that the topology inherently compensates for any resonance mis-tuning arising from practical non-idealities - through counterbalancing effects during the dead time. This work investigates and models this physical phenomenon in detail, and predicts the minimum dead time required for stable operation as a function of resonance mis-tuning. These developed models would be useful for designing an SRC to operate as a bidirectional, uncontrolled DCX that is tolerant to resonance mistuning.

ABSTRACT. An analytical model of the bidirectional AC-AC Dual Active Bridge (DAB) converter is developed. Since the AC-AC DAB converter has grid, switching, and side-band harmonics, it cannot be modeled using the traditional Generalized Average Method (GAM), which deals only with one frequency and its harmonics. A proposed sixteenth order continuous-time EGAM model that considers the aforementioned harmonics is derived. Validation again PLECS has been provided to highlight model accuracy.

ABSTRACT. In this article, the design and characterization of a 1200 V, 200 A rated printed circuit board (PCB) embedded silicon carbide (SiC) MOSFET half-bridge are being presented. The layout of the power module has been optimized to reduce the power loop inductance, achieving a 2.1 nH of power loop inductance. The dimensions of the fabricated module are 48 mm x 32 mm (L x B), with thickness being 0.35 mm, leading to a reduction of 57% in footprint and a reduction of 95% in volume compared to commercially available half-bridge SiC MOSFET power modules of 1200 V, 200 A with GM4 package.

ABSTRACT. Realizing modular multilevel versions of single-stage converters for single-phase AC to DC applications is particularly challenging due to the need to manage twice the line-frequency power. This power must be buffered by internal energy storage elements, which can lead to significant energy storage requirements being imposed on the converter. An additional challenge is ensuring smooth transitions when the grid voltage polarity reverses every half cycle. In this digest, a novel self-balancing single-stage modular multilevel converter is proposed that can scale based on the voltage and therefore is suitable for distribution-level applications. This structure leverages submodules that feature two half-bridges connected in an anti-series arrangement. The converter is operated such that the submodule dc-link voltages naturally follow the grid voltage. The topology shares a similar energy transfer mechanism as the dual active bridge converter for DC-DC applications. This digest introduces the topology, discusses key operating principles and provides transient simulation results for a representative application. The full paper will provide experimental results using a laboratory-scale prototype.

ABSTRACT. This research focuses on the development of an ultra-compact and high-efficiency Quad-Active-bridge (QAB) DC-DC multiport converter, which can be used to interconnect PV panels, battery energy storage (BES), and electric vehicles (EVs) with a DC distribution grid. The emerging Gallium Nitride (GaN) switching devices are utilized to achieve high switching frequency, higher efficiency, and high power density. Additionally, a multi-winding planar transformer is employed in the QAB converter to provide galvanic isolation and couple different power ports for flexible power management. Single-phase shift (SPS), Extended-phase shift (EPS), and Dual-phase shift (DPS) modulation strategies are investigated and compared to regulate the power flow in various operating modes. Electro-thermal modeling and simulation of the QAB converter are conducted, and a 15kW GaN-based QAB converter prototype is developed for experimental validation.

ABSTRACT. With the growing number of retired electric vehicle batteries, the second-life battery energy storage system (2-BESS) has attracted significant research attention due to its cost-effectiveness and environmental benefits. This paper proposes a battery-integrated cascaded H-bridge (BI-CHB) 2-BESS architecture equipped by a hierarchical partial power processing (HiPPP) network, effectively addressing the challenges posed by second-life battery (SLB) heterogeneity. The proposed HiPPP network is designed based on a novel statistical framework integrating multivariate distribution flattening and mixed-integer linear programming (MILP). The study reveals a new trade-off frontier among active power capability, reactive power capability, and energy utilization. Importantly, the HiPPP network significantly improves the power-energy trade-off and demonstrates robust reactive power injection capabilities, essential for voltage support in grids with high renewable penetration. Moreover, the HiPPP network effectively mitigates the adverse effects of SLB heterogeneity, enhancing reliability and reducing uncertainty in the output power of 2-BESSs.

ABSTRACT. Oceans possess many forms of renewable energies including thermal, tidal, and wave energy. There is a significant potential to utilize marine energy resources. In the United States, the total amount of marine energy available is equivalent to about 57% of the country's total power generation in 2019. Even if a fraction of this technical potential is harnessed, marine energy technologies could play a crucial role in fulfilling the nation's energy requirements. However, ocean environments present many challenges for cost-effective renewable energy conversion, including optimal control of an ocean wave energy. This report presents a novel cost-effective energy conversion technique for ocean wave energy harvesters. The proposed system is simulated by using the Oak Ridge Converter to directly interface ocean wave energy source with the grid. The system description and simulation results are presented. The results show the proposed system is a promising technology to reduce infrastructure costs for ocean wave energy harvesters.

ABSTRACT. Modular Multilevel Converters (MMC) have gained popularity for high power integration. In the same domain of applications, energy storage (ES) integrated MMC system is used to provide grid ancillary services involving voltage and frequency regulation. Offshore wind power has made a major contribution to the grid renewable power generation. The grid integration of offshore wind energy poses two major challenges. One of them is the requirement of voltage-frequency regulation at point of common coupling (PCC) and secondly, harvesting active power to compensate for fluctuations in wind power. The ES-STATCOM based MMC system caters to these challenges and enhances grid reliability by enabling active power compensation for peak loads. This paper defines the stable operating region of ES-STATCOM and based on that proposes grid forming control to integrate offshore wind power. The proposed control architecture is validated in a test system environment through a Real-Time Digital Simulator (RTDS) and Virtex 7-based FPGA controller in a Control Hardware-in-Loop (C-HIL) environment.

ABSTRACT. The dual comparison one cycle control (DC-OCC) proposed in this paper regulates both the peak and valley of the current, unlike conventional one cycle control (C-OCC), which controls only the peak. This dual control approach effectively eliminates the steady-state DC offset, current distortion, and localized sub-harmonic instability issues observed in C-OCC. The DC-OCC structure is further extended to support bidirectional power flow, adjustable power factor operation, and STATCOM functionality. These enhancements are achieved with relatively simple modifications compared to existing methods in the literature. An average model of the DC-OCC is developed to analyze instability when the converter draws or injects lagging current into the grid. The controller's effectiveness is validated through detailed simulation and experimental studies.

ABSTRACT. In this paper, the high-frequency model of PEBB 6000, a 6 kV, 1 MW full-bridge power electronics building block (PEBB) incorporating 10 kV SiC MOSFET-based power modules is developed as a perfect representative of medium-voltage, high-power converters. The developed model primarily relies on measurement but also utilizes finite element analysis (FEA) when a measurement is not applicable. The power stage is modeled, focusing on switching transient and noise propagation path modeling. Power modules are characterized and nonlinearly modeled in Saber RD; module-to-heatsink parasitic capacitance, DC busbar, capacitor daughtercards, heatsink-to-fan array parasitic capacitance, gate drivers, and isolated auxiliary power system (APS) power supplies complete circuit models along with their parasitic capacitances are measured and added to the model, focusing on the wireless power transfer (WPT) and current-transformer based gate-driver power supply (GDPS), as discussed throughout the digest. The study focuses on the impact of these components on the noise propagation paths and how they are impacted from the noise in one unit and neighboring units in a modular multi-level converter (MMC). The developed high-frequency model facilitates the examination of noise propagation paths under load conditions and the crosstalk between PEBB units in MMCs, enabling the assessment of the influence of neighboring units on others.

ABSTRACT. The rapid expansion of AI-driven data centers is increasing the demand for more efficient high-power isolated ac/dc conversion solutions. Conventional matrix-type isolated-single-phase-HF-link three-phase (3-Φ) converters face significant challenges in high-power applications due to excessive transformer current stresses. This digest proposes a novel isolated-three-phase-HF-link matrix-type three-phase ac/dc converter (i3X-Rectifier) that enhances power efficiency and volumetric density in high power, i.e., 100 kW range, data center power supplies. The i3X-Rectifier incorporates a 3-Φ isolation transformer and employs either an indirect-matrix or direct-matrix front-end, enabling direct single-stage conversion from 3-Φ low-frequency (LF) mains to 3-Φ high-frequency (HF) transformer voltages. Key advantages of the proposed i3X-Rectifier include significantly reduced rms/peak current stress on power transistors and transformer windings, lower flux amplitude in the magnetic core, improved core utilization, and minimized input/output filtering requirements. These benefits align with the performance advancements of 3-Φ dual-active-bridge (DAB) dc/dc converters over their 1-Φ DAB counterparts, positioning the i3X-Rectifier as a competitive solution for high-power applications. Theoretical analysis of the operating principle is validated through closed-loop circuit simulations, with an outlook to be presented in the final paper.

ABSTRACT. This digest presents a comprehensive design and optimization methodology for a novel GaN-based transformerless online uninterruptible power supply (UPS) topology that enables universal input/output operation with a common neutral and a single DC bus. The topology, featuring multi-mode operation, allows the bus voltage to be lower than the peak line voltage, enabling the use of commercially available 650-V GaN devices. This enhances switching performance, reduces system size, and improves battery integration. A design tool using the proposed system-level design methodology is developed to sweep key variables—including bus voltage, switching frequency, GaN device, number of parallel devices, inductor values, core material/shape, and heatsinks—while computing loss, volume, and cost for each design point. Pareto-optimal fronts are used to identify the best trade-offs amongst loss, volume, and cost. A 1.2-kVA prototype was designed to achieve 96% ac-ac efficiency. For experimental validation, a GaN-based prototype was built and tested under stable closed-loop control. The developed methodology for the first-of-its-kind topology enhances UPS performance, and enables efficient, data-driven UPS design tailored to practical performance, size, and cost constraints.

ABSTRACT. Accurate determination of switching loss is essential for soft-switched converters, such as dual active bridge (DAB) converters, operating at high switching frequencies. The established analytical methods become complex in the presence of nonlinear device capacitors and layout and device parasitics. This paper introduces an experimental method to characterize switching losses, including turn-off and partial zero-voltage switching (ZVS) turn-on losses, for different operating modes of DAB using an actual converter setup. The proposed method is validated through SPICE simulations and experiments conducted on a 10kW SiC-based DAB hardware prototype.